44th Lunar and Planetary Science Conference (2013)
2819.pdf
SEASONAL RELEASE OF WATER VAPOR BY GROUND ICE ON MARS: IMPLICATIONS FOR
SURFACE FROSTS AND ATMOSPHERIC WATER ABUNDANCE. Jonathan Bapst and Stephen E. Wood,
Department of Earth and Space Sciences, University of Washington, Box 351310, Seattle, WA 98195.
([email protected])
Introduction: Observations from the Mars Odyssey Neutron Spectrometer (MONS) suggest substantial
water ice exists beneath the Martian surface (in the
upper meter or so) extending from the polar regions to
mid-latitudes in both hemispheres [1,2]. The boundary
between regions with ground ice and those without
appears to be abrupt, as predicted [3].
Coupled thermal and vapor diffusion models have
been used to study the behavior of water vapor as it
diffuses through Martian regolith on intra-annual timescales [3,4]. These models show that when conditions
are favorable (i.e. adequate atmospheric water abundance and sufficiently cold subsurface temperatures)
pore ice will form from diffused water vapor.
Common in these models, surface vapor density is
treated as a constant value which will drop to saturation vapor pressure when the surface cools beneath the
frost point temperature (e.g. at night or during the winter season). When this happens there can exist a vapor
density gradient such that favors transport of water
vapor from the subsurface to the surface.
Our work here is aimed at estimating the amount of
vapor that can be transported during seasonal transitions. We will explore various parameters (e.g. depth
to perennial ice, atmospheric water abundance, geographic properties, etc.) that can affect the outward
flux of vapor and its timing.
Vapor that is transported to the surface as described
above would condense as frost if conditions are right
(i.e. cold surface temperatures and lower vapor density
at the surface than the near subsurface). The ultimate
fate of that frost is sublimation and mixed into the atmosphere. This process, because of the widespread
occurrence of ground ice on Mars, has the potential to
influence the season water cycle on Mars.
Methodology: In order to estimate the vapor flux
from the subsurface to the surface (in order to form
frost and/or be mixed into the atmosphere) we focus on
the upper grid points in our 1-D finite difference numerical model (similar to those described [3,4]) which
allows for pore ice formation.
Vapor flux
is calculated based on the diffusivity
of the regolith
and the vapor density gradient
( ) via Fick’s Law:
and are the porosity and tortuosity of the regolith, respectively.
We calculate our gradient using the surface vapor
density and the vapor density just below the surface
(i.e. 2nd grid point; this should capture transport between the surface and subsurface).
,
,
,
,
assume
4 10"# $ &% after [5],
0.40 and
3. The location of z2 is ~0.5 cm
' (
below the surface. We calculate diffusion of vapor and
temperature along with ice adjustment 1000 times per
sol. We ignore CO2 frost and the effect of adsorbed
water but recognize these could dramatically alter our
results (see [4] for treatment of adsorbed water).
Preliminary Results: We ran our vapor diffusion
model for a general case at 45°N using a prescribed
atmospheric water abundance of 1 Pa (to ensure stable
ground ice) and a surface albedo of 0.15. We start with
an initially dry soil but conditions are such that perennial ice is stable below ~1 m depth. Due to the large
fluctuations in surface vapor density our flux sign
changes on a diurnal basis with vapor moving in and
out of the upper subsurface layers.
We
Figure 1. Vapor flux out of the subsurface (positive) and into
the subsurface (negative). Variation in surface vapor density
due to nighttime cooling produces changes in the sign of the
flux on diurnal bases.
Due to the rapid variation in surface vapor density,
it is useful to smooth out of the data (averaging over
each Martian sol) to see seasonal variation. By subsequently integrating over each sol we can approximate
inward/outward mass transport of vapor per sol (i.e.
resetting each sol; Figure 2). We convert from kg/m2 to
44th Lunar and Planetary Science Conference (2013)
precipitable microns for comparison to observations of
column-integrated atmospheric water vapor [6]. This
might also represent maximum frost thicknesses if
surface frost comes and goes on a diurnal basis [7].
The sharp change at Ls ≈ 340° is an artifact of the
model, in reality we expect this to a smoother transition as seasonal pore ice sublimates and diffuses from
the near surface (and thus the steep near-surface vapor
density gradients diminish).
Figure 2. The integrated transport of vapor per Martian sol
(using daily-averaged vapor density gradient). Note the
anomalous, transient release of vapor to the surface (positive)
around Ls ≈ 330°. Integrated over that time period accumulates to 20+ pr-µm transported to the surface/atmosphere
system.
Discussion: The release of vapor during late winter
is an interesting result. Our explanation for this is the
late-winter warming of near-surface seasonal ice, generating steep vapor gradients and thus transporting
water towards the surface.
Figure 3. Near-surface seasonal ice shows growth and depletion in late winter/early summer. The depletion of ice at 0.5
cm depth (used for water vapor flux calculation of above)
corresponds to the sharp decrease in flux around Ls = ~340°
(see Figure 2).
Seasonal Water Frost: The possible addition of
tens of precipital microns to a cold, dry winter atmosphere on Mars could result in significant surface frost.
Smith et al., 2002 show northern winter atmospheric
vapor abundance in the mid-latitudes (and above) to be
below the detection threshold. More work needs to be
done, including atmospheric modeling, in order to de-
2819.pdf
termine the holding capacity and mixing ability of the
atmosphere with respect to water vapor on these timescales. Spacecraft observations will be important in
testing our model’s predictions.
Apart from orbiting spacecraft, the Viking Lander
2 observed surface water frost at ~48°N [7]. Here frost
was observed mostly in the late winter season. Some of
this frost may have been sourced from probable ground
ice in the region.
The ability for ground ice to release vapor provides
a mechanism for local changes in albedo due to subsurface volatile content. This could play interesting roles
in the dynamic of Martian climates, if such changes
(e.g. obliquity) can strongly affect the process described here.
Seasonal Atmospheric Water Cycle: The seasonal
water cycle of Mars has been observed by multiple
instruments [8,9] and is relatively well replicated by
general circulation models [14]. The addition of tens of
microns of water into the atmosphere from ground ice
sources could have a significant impact on the global
water budget. Smith et al., 2002 mentions a large increase in water vapor in the mid-to-high northern latitudes of Mars prior to summer solstice. This could be
one possible explanation not yet considered.
References: [1] Feldman, W.C. et al. (2002) Science, 297, 75. [2] Boynton, W. V. et al. (2002) Science, 297, 81. [3] Mellon, M. T. and Jakosky, B. M.
(1993) JGR, 98(E2). [4] Schorghofer, N. and Aharonson, O. (2005) JGR, 110, E05003. [5] Hudson, T. L.
et al. (2007) JGR, 112, E05016. [6] Smith, M. D.
(2002) JGR, 107, E11. [7] Svitek, T. and Murray, B.
(1990) JGR, 95, B2. [8] Jakosky, B. M. and Farmer, C.
B. (1982) JGR, 87, 2999-3019. [9] Smith, M. D.
(2002) JGR, 107, E11. [10] Madeleine, J.-B. et al.
(2009) Icarus, 203(2).